WARNING:tensorflow:Reduce LR on plateau conditioned on metric `val_loss` which is not available.
WARNING:tensorflow:Gradients do not exist for variables when minimizing the loss.ġ/660 - ETA: 8:05:51WARNING:tensorflow:Can save best model only with val_accuracy available, skipping. Train for 660 steps, validate on 1024 samples WARNING:tensorflow:sample_weight modes were coerced from Please use Model.fit, which supports generators. WARNING:tensorflow:From :6: Model.fit_generator (from .training) is deprecated and will be removed in a future version. When I run the model fit I'll get this warning & error message.
Global_average_pooling2d_5 ( (None, 1024) 0īatch_normalization_10 (Batc (None, 2048) 8192īatch_normalization_11 (Batc (None, 512) 2048Įmbedding_1 (Embedding) (None, 13, 100) 1300Īdding embedd_layer and flatten layer after the layer prediction resulted in this error to pop-up, pile(optimizer='adam', loss='binary_crossentropy', metrics= METRICSįeature (InputLayer) 0 Model = (inputs= input_x, outputs = final_predictions) Predictions = Dense(len(all_labels), activation='relu')(dropout2_x)Įmbedd_layer = Embedding(13, 100, weights=, trainable=False,name = "embedding_1")(predictions)įlatten = Flatten(name="flatten")(embedd_layer)įinal_predictions = Dense(13, activation='relu')(flatten) Input_x = Input(shape = (img_input.shape), name='feature')īase_model = MobileNet(include_top= False, input_shape=img_input.shape, weights='imagenet')(input_x)Īvgpool_x = GlobalAveragePooling2D()(base_model)ĭense1_x = Dense(2048, activation='relu')(avgpool_x)īatch1_x = BatchNormalization()(dense1_x)ĭense2_x = Dense(512, activation='relu')(dropout1_x)īatch2_x = BatchNormalization()(dense2_x) Ntdll.Def build_mobilenet_embedd_model(img_input ): ShooterGame.exe!GuardedMainWrapper() (0x000000013fdf32ca) + 5 bytes ShooterGame.exe!FD3D11DynamicRHI::RHICreateUniformBuffer() (0x00000001405477ca) + 0 bytes Assertion failed: MappedSubresource.RowPitch >= NumBytes
0 kommentar(er)
